Wind Energy Time-Series Analysis

Wind energy datasets are large and time-indexed, with turbine output tied closely to weather conditions over time. Getting useful analysis out of that requires handling the data volume efficiently before any modelling question can even be asked properly.

Built a processing pipeline in Python and Pandas to handle the dataset at scale, then applied time-series analysis techniques to look at patterns in output over time and its relationship to weather variables, with the heavier processing run on Azure.

Pandas for large-scale data wrangling, Azure for compute where local processing wasn't practical, time-series decomposition and visualization to surface patterns in turbine output.

TODO: add the specific findings — which patterns/correlations the time-series analysis surfaced, and any visualizations worth including here.

Data engineering and data science aren't separate phases in practice — most of the interesting analytical decisions here were shaped by how the data had to be processed first.